creep, fracture, damage parameter, neural networks, parameter identification, systems of ordinary differential equations.

The application of neural network modeling methods to solve the creep and fracture model parameters identification problem of anisotropic titanium alloy 3V specimens described by energy variant of creep theory system of differential equations is considered. The process of neural network solution constructing for this problem is developed. By using of this process the model parameters are found. Solution corresponding to these parameters is in good agreement with the theoretical deformation-strength characteristics, the experiment and the results of other authors.

1. Vasilyev Alexander Nikolaevich Dr. Sci. Tech., Professor, St. Petersburg state polytechnic university (national research university), St. Petersburg 2. Kuznetsov Evgenii Borisovich Dr. Sci. Phys. & Math., Professor, Moscow aviation institute (national research university), Moscow 3. Leonov Sergey Sergeevich Postgraduate student, Moscow aviation institute (national research university), Moscow